review-agents-md
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to perform file system operations using shell commands such as
ln -sandmvto organize and symlink project configuration files (AGENTS.md,CLAUDE.md). - [INDIRECT_PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by reading and refactoring existing documentation files from the user's project, which may contain adversarial instructions.
- Ingestion points: The agent reads existing
AGENTS.mdandCLAUDE.mdfiles from the project root during Phase 0 and Phase 1. - Boundary markers: Absent. The instructions do not define delimiters or provide specific prompts to ignore instructions within the analyzed data.
- Capability inventory: The skill possesses the capability to read and write files, rename files (
mv), and create symbolic links (ln -s). - Sanitization: Absent. No sanitization, escaping, or validation logic is applied to the content retrieved from existing project files before it is processed or merged into new versions.
Audit Metadata